Collaborative notes https://docs.google.com/document/d/1ENavslbeGXdWHFb8uY8qVqHDOmELUeOe/edit

This session explores how the EOSC-Future/RDA Artificial Intelligence and Data Visitation Working Group (AIDV-WG) outputs can be applied to implement the key principles and recommendations found in global AI governance frameworks, including those issued by the G20, United Nations, African Union, EU, Latin America, USA, International Science Council (ISC), and the Royal Society.
The session will demonstrate how the AIDV-WG’s practical tools – such as the global adaptation of the AI Bill of Rights, consent guidance for data visitation, secure processing environment protocols, and ethics review criteria – can support responsible and coordinated implementation of international AI policy in scientific research environments.
Specific objectives of the session are to:
1. Demonstrate how AIDV-WG outputs support the operational goals of global AI governance frameworks, with particular attention to the G20’s call for ‘AI for Good and for All,’ the UN Pact for the Future, and the African Union’s Continental AI Strategy.
2. Showcase practical applications of AIDV tools in research settings, including cross-border and federated AI research projects that require data access, ethical oversight, and regulatory compliance.
3. Identify challenges and solutions in securing informed consent and ethical oversight for data visitation, especially in AI research involving sensitive data or complex data flows.
4. Discuss implementation strategies for secure processing environments (SPEs) that align with the safety, transparency, and accountability requirements set by international frameworks.
5. Propose steps to align national and institutional research practices with high-level AI governance priorities, focusing on interoperability, policy coherence, and reproducibility in AI-driven scientific research.
6. Encourage international coordination on the application of AIDV-WG tools, especially among countries and institutions building foundational capacity for AI governance.
7. Foster engagement with science policy bodies and multilateral institutions to explore how AIDV-WG deliverables can inform the development of harmonized guidelines, capacity-building programs, and technical standards for responsible AI in research.

90-Minute Session Agenda

• 0:00 – 0:10 (10 min) | Welcome and Session Framing
  Brief introduction to the EOSC-Future/RDA AIDV Working Group and its core outputs.
  • Overview of global AI governance developments (G20, UN Pact for the Future, ISC, AU, Latin America, EU, Royal Society).
  –> Framing question: How can we move from international AI principles to real-world research practice?
• 0:10 – 0:30 (30 min) | Use Case Presentations (3 Brief Cases)
  • Three short case presentations (~6 minutes each + transition) covering:
  o Application of AIDV consent guidance in federated health data research
  o Implementation of secure processing environments aligned with policy standards
  o Use of AIDV ethical review guidance in cross-jurisdictional scientific AI projects
  –> Each use case demonstrates alignment with at least one international framework (e.g., G20, UN, AU).
• 0:30 – 0:45 (10 min) | Interactive Poll + Open Floor Reaction
  –> Live poll or mentimeter to gauge participant perspectives on challenges and readiness (e.g., “Which AIDV output is most immediately applicable in your context?”)
  • Open mic or chat-based response to poll results
  • Guided reflections on institutional, regional, and technical barriers to implementation
• 0:45 – 1:00 (15 min) | Structured Small Group Discussion (Breakout)
  • Participants divide into small groups (5–7 people) to address one of the following prompts:
  1. How can AIDV tools support national AI strategies or policy development?
  2. What’s needed to apply AIDV secure processing guidance in local infrastructure?
  3. What would an effective institutional or regional adoption roadmap look like?
  –> Each group designates a rapporteur to capture key points.
• 1:00 – 1:15 (10 min) | Group Reports and Comparative Reflections
  • Each group rapporteur shares key insights (2–3 minutes per group)
  –> Facilitator summarizes overlapping themes, differences, and regional considerations
• 1:15 – 1:25 (10 min) | Roadmapping: Priorities and Next Steps
  • Facilitator presents a draft set of next steps for AIDV alignment with G20, UN, AU, EU, and ISC frameworks
  • Participants provide input to refine proposals for new deliverables, pilots, or training activities
  • Poll to prioritize implementation actions (e.g., toolkit development, training modules, guidance alignment)
• 1:25 – 1:30 (5 min) | Closing Summary and Call to Action
  • Summary of outcomes and insights
  –> Next steps

This session is designed for members of the Research Data Alliance (RDA) and International Data Week (IDW) communities who are actively engaged in the development, implementation, or governance of AI-enabled research and data infrastructures. The target audience includes data scientists, research software engineers, infrastructure and repository managers, ethics committee members, policymakers, and institutional leaders responsible for research integrity and data governance. It will also appeal to participants from national science agencies, international organizations, and global research collaborations seeking to align with evolving AI governance frameworks such as those from the G20, United Nations, African Union, European Union, and International Science Council. Given the Plenary 25 focus on “Data for Positive Change,” the session invites those interested in translating high-level policy into practice, particularly across federated, cross-border, and multidisciplinary data environments.